Presentation is loading. Please wait.

Presentation is loading. Please wait.

DNAseq analysis Bioinformatics Analysis Team

Similar presentations

Presentation on theme: "DNAseq analysis Bioinformatics Analysis Team"— Presentation transcript:

1 DNAseq analysis Bioinformatics Analysis Team
McGill University and Genome Quebec Innovation Center

2 Module #: Title of Module

3 What is DNAseq ? DNA sequencing is the process of determining the precise order of nucleotides within a DNA molecule. The advent of rapid DNA sequencing methods has greatly accelerated biological and medical research and discovery.

4 Why dnaSeq ? Whole genome sequencing: Whole exome sequencing:
Whole genome SNV detection Structural variant Capture the regulatory region information Cancer analysis De novo genome assembly Whole exome sequencing: Cheaper Capture the coding region information Rare diseases analysis

5 What the DNAseq problem is about ?
Strings of 100 to ≈1kb letters Puzzle of 3,000,000,000 letters Usually have 120,000,000,000 letters you need to fit Many pieces don’t fit : sequencing error/SNP/Structural variant Many pieces fit in many places: Low complexity region/microsatellite/repeat

6 DNAseq overview

7 DNAseq: Input Data

8 Input Data: FASTQ End 1 End 2 Sample1_R2.fastq.gz Sample1_R1.fastq.gz
Each sample will generate between 5Gb (100x WES) to 300Gb (100x WGS) of data

9 Q = -10 log_10 (p) Where Q is the quality and p is the probability of the base being incorrect.

10 QC of raw sequences

11 QC of raw sequences low qualtity bases can bias subsequent anlaysis
(i.e, SNP and SV calling, …)

12 QC of raw sequences Positional Base-Content

13 QC of raw sequences

14 QC of raw sequences Species composition (via BLAST)

15 DNA-Seq: Trimming and aligning

16 Read Filtering Clip Illumina adapters: Trim trailing quality < 30
Filter for read length ≥ 32 bp

17 Assembly vs. Mapping mapping all vs reference Reference reads contig1
all vs all

18 RNA-seq: Assembly vs Mapping
Reference-based mapping Ref. Genome DNA-seq reads contig1 contig2 De novo assembly

19 Read Mapping Mapping problem is challenging:
Need to map millions of short reads to a genome Genome = text with billons of letters Many mapping locations possible NOT exact matching: sequencing errors and biological variants (substitutions, insertions, deletions, splicing) Clever use of the Burrows-Wheeler Transform increases speed and reduces memory footprint Used mapper: BWA Other mappers: Bowtie, STAR, GEM, etc.

20 SAM/BAM Used to store alignments SAM = text, BAM = binary
Sample1.bam between 10Gb ot 500Gb each bam Sample2.bam SRR M = NCCAGCAGCCATAACTGGAATGGGAAATAAACACTATGTTCAAAG Used to store alignments SAM = text, BAM = binary SRR M = NCCAGCAGCCATAACTGGAATGGGAAATAAACACTATGTTCAAAGCAGA Read name Flag Reference Position CIGAR Mate Position Bases Base Qualities SAM: Sequence Alignment/Map format

21 Sort, View, Index, Statistics, Etc.
The BAM/SAM format Sort, View, Index, Statistics, Etc. $ samtools flagstat C1.bam in total (QC-passed reads + QC-failed reads) 0 + 0 duplicates mapped (100.00%:nan%) paired in sequencing read1 read2 properly paired (85.06%:nan%) with itself and mate mapped singletons (3.44%:nan%) with mate mapped to a different chr with mate mapped to a different chr (mapQ>=5) $

22 DNA-Seq: metrics

23 Included metrics Metrics are collected form the output of Trimmomatic, Samtools and Picard softwares

24 DNA-Seq: Alignment refinement

25 Local indel realignment
Primary alignment with BWA Local re-alignment around indels with GATK Possible mate inconsistency are fixed using Fixmate image:

26 Duplicates and recalibration
Mark duplicates with Picard Base Quality Score Recalibration GATK Example Bias in the qualities reported depending of the nucleotide context image:

27 DNA-Seq: SNV calling

28 Single Nucleotide Variant calling
Aim: differentiate real SNPs from sequencing errors An accurate SNP discovery is closely linked with a good base quality and a sufficient depth of coverage sequencing errors SNP

29 SNP and genotype calling workflow
Variants from multiple samples are called simultaneously using the mpileUp method from samtools and quality filtered using bcftools Bayesian apporachs MLE apporachs Nielsen et al June 2011

30 The variant format : vcf
Variant Call Format Column FORMAT defines “:” separated values GT = Genotype DP = depth

31 VCF visualization in IGV

32 DNA-Seq: SNV annotation and metrics

33 Variant annotation Hypo- or hyper-mappabilty flag dbSNP [SnpSift]
Mark SNV in low confidence regions dbSNP [SnpSift] Mark already known variant Variant effects [SnpEff] predict the effects of variants on genes (such as amino acid changes) dbNSFP [SnpSift] Functional annotations of the change Cosmic[SnpSift] Known somatic mutations

34 SNV statistics Statistics are generated from the SNPeff stats outputs
Example of one of the SNv metrics graph

35 DNA-Seq: Generate report

36 Home-made Rscript Generate report Files generated:
Noozle-based html report which describe the entire analysis and provide QC, summary statistics as well as the entire set of results Files generated: index.html, links to detailed statistics and plots For examples of report generated while using our pipeline please visit our website

Download ppt "DNAseq analysis Bioinformatics Analysis Team"

Similar presentations

Ads by Google